INTRODUCTION
A numerical description of rock joint roughness is necessary to describe adequately the pseudostatic and dynamic rock-joint behavior. The most commonly used measure of joint roughness in rock engineering practice is the joint roughness coefficient (JRC), proposed by Barton [1] and adopted by the International Society for Rock Mechanics [2]. Three basic methods for the determination of JRC have been proposed by various investigators - a tilt test on rock joints as proposed by Barton and Choubey [3], using an average deviation from the center line about which roughness is measured [4,5], and using the theory of fractal geometry.
These methods are evaluated in this paper to determine their appropriateness for use in the assessment of rock-joint behavior of Apache Leap tuff joints by comparing the estimates using these methods to the JRC values back-calculated from the laboratory direct shear test results.
DETERMINATION OF JOINT ROUGHNESS COEFFICIENT
As proposed by Barton and Choubey [3], the tilt test involves a determination of the tilt angle at which the top block of a mated joint specimen begins to slide along the proposed direction of shear. Once a tilt angle is obtained, the JRC can be calculated using the following equation:
[Equation available in full paper] (1)
where á is the tilt angle, Ör is the joint residual angle of friction, JCS is the joint wall compressire strength, and óho is the corresponding effective normal stress calculated from the weight of the top block of the joint specimen when sliding occurs. Ör and JCS in Eq. (1) need to be determined separately and, consequently, may introduce some uncertainties into the determination of JRC and can be found elsewhere [3].
